Output details

Probabilistic Constrained MPC for Multiplicative and Additive Stochastic Uncertainty

For systems with stochastic parameters and unknown disturbances, this paper proposes a method of imposing system constraints with prescribed probabilities that prevents the constrained control problem from becoming infeasible at a later time. This allows a robust stability analysis to be applied to stochastic model predictive control for a wide class of uncertain systems. The problem definition and computational framework introduced in this paper are the basis of recent developments in stochastic model predictive control with guaranteed robustness. This work has led to a collaboration with the British Antarctic Survey on ecosystem-based management of fisheries (contact details available).